ALERT! Warning: your browser isn't supported. Please install a modern one, like Firefox, Opera, Safari, Chrome or the latest Internet Explorer. Thank you!
Log inLog in to ACORN Wiki. | RegisterSign on to ACORN Wiki.

Pre-post change score

The pre-post change score is the difference between the test score at the beginning of treatment and the score at the final assessment. By convention, the score is calculated so that a positive value indicates change in the desired direction. When evaluating change in real world treatment settings, the intention-to-treat method is generally preferred when determining the post assessment.

The pre-post change score has the advantage of being an intuitively obvious method of measuring outcome, and is simple to calculate. So long as the user is familiar with the test and the range of possible scores, the change score provides a face valid measure of improvement.

Test scores will very from one administration to another due to imperfect reliability. In order to test if the change in scores is greater than would be expected from random variation alone, many researcher's advocate the use of the Reliable Change Index.

Raw change scores are often converted to effect size (a standardized change score), in order to permit comparison of results across different samples using different outcome measures.

However, simple change scores have a serious drawback as a measure of outcome: the change score is strongly correlated with the intake score. Patients reporting high levels of distress will tend to have larger change scores at the end of treatment than patients reporting milder symptoms. See the topic RegressionToMean for more on this phenomena.

This means that it is impossible to determine whether one patient achieved a better outcome than another by comparing simple change scores with out first knowing if their intake scores were comparable. Likewise, comparing average change scores for one provider to those of another provider is meaningless because the providers may differ significantly in the mix of cases being treated.

In order to meaningfully compare outcomes from one group of patients to another, it is necessary to utilize a case mix adjustment model. Typically this is done through use of the generalized linear model, which permits the calculation of a residualized change score.

-- JebBrown - 07 Jan 2007
This site is powered by FoswikiCopyright © by the contributing authors. All material on this collaboration platform is the property of the contributing authors.
Ideas, requests, problems regarding ACORN Wiki? Send feedback
Syndicate this site RSS
^